A Masters student from Robert Gordon University, Aberdeen, Scotland, has found a way to improve the performance of one of the worlds leading search engines.
Using the spare cycles of the computers at the University, Maryam Kontagora from Abuja, Nigeria, who has been studying at the School of Computing, ran a myriad of concurrent MapReduce jobs to simulate hundreds of users searching and sorting on Google at the same time, ultimately improving the results of these searches.
The MSc student has just been informed that an article based on the findings she has put together for her Masters dissertation has been accepted at a top international conference. The IEEE International Conference on Complex, Intelligent and Software Intensive Systems (CISIS), to be held at Krakow, in February 2010 will give Maryam a forum in which to showcase her findings.Her work describes how to improve benchmarks of the search engine's MapReduce environment on multiple virtual machines. According to the international reviewers for the CISIS conference, her article 'is certainly interesting for the audience and a very good and solid work'.
Introduced by Google in 2004, MapReduce is a programming environment to allow hundreds of computers to work together on gaining results for complicated searches. Thousands of MapReduce jobs are executed on Google's clusters every day, meaning that lots of the searches we run on our favourite restaurants and TV shows will be executed by using a MapReduce tool.
Maryam explained, "My ultimate career goal is to help advance knowledge in the field of parallel computing and how it may be used to improve the performance of large tasks, like searches that would normally take a long time to execute."
Dr Horacio Gonzalez-Velez, a lecturer with the School of Computing who is Maryam's academic supervisor for her dissertation, said, "Maryam has always demonstrated an outstanding ability for computing. She is a very clever young woman with an ability to identify useful connections among concepts that other people miss. Her work is definitely a step in the right direction and will hopefully help us to gain further insights into the performance of parallel computers."
Professor Ian Allison, Head of the School of Computing added, "This project is an excellent example of how our applied research shapes our teaching. Students get to work with leading edge technologies and ideas. I'm delighted that Maryam will have a chance to discuss her work to an international audience."